there are many pop sci articles and press releases wrt quantum computing these days and some may tend to stretch precision in description. one particular area that bothers me at the moment: some articles such as this statement from IBM  (and there are many other typical cases/ examples, maybe even some in papers/ books etc) say things in the form:
With a quantum computer built of just 50 qubits, none of today’s TOP500 supercomputers could successfully emulate it, reflecting the tremendous potential of this technology.
many researchers have been quoted making similar statements and they seem to be rarely backed by citations or details. now ofc this relates to the purity of the qubits and presumably claims like this are skipping the calculations that decrease capability based on error correction consuming some of the "qubit capability."
what is the scientific basis/ details for claims of the form, "x qubits could exceed classical computers of capability/ size y"?
now, am not at all questioning the general concept here shown decades ago that QM computers can in principle duplicate or outperform classical computers etc. but in other words, is there a rough estimate for # of qubits vs classical processor speed etc., what papers consider this question, what are the details of that calculation/ estimate? for answers, am willing to forego estimates of error correction, although would regard that as a bonus if available.